[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20220829141517.bcjbdk5zb74mrhgu@wittgenstein>
Date: Mon, 29 Aug 2022 16:15:17 +0200
From: Christian Brauner <brauner@...nel.org>
To: Alexey Khoroshilov <khoroshilov@...ras.ru>
Cc: Alexander Viro <viro@...iv.linux.org.uk>,
Rustam Subkhankulov <subkhankulov@...ras.ru>,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
ldv-project@...uxtesting.org
Subject: Re: [PATCH] fs/inode.c: change the order of initialization in
inode_init_always()
On Sat, Aug 27, 2022 at 05:09:26PM +0300, Alexey Khoroshilov wrote:
> From: Rustam Subkhankulov <subkhankulov@...ras.ru>
>
> If function 'security_inode_alloc()' returns a nonzero value at
> [fs/inode.c: 195] due to an error (e.g. fail to allocate memory),
> then some of the fields, including 'i_private', will not be
> initialized.
>
> After that, if the fs-specfic free_inode function is called in
> 'i_callback', the nonzero value of 'i_private' field can be interpreted
> as initialized. As a result, this can cause dereferencing of random
> value pointer (e.g. nilfs2).
>
> In earlier versions, a similar situation could occur with the 'u' union
> in 'inode' structure.
>
> Found by Linux Verification Center (linuxtesting.org) with syzkaller.
>
> Signed-off-by: Rustam Subkhankulov <subkhankulov@...ras.ru>
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> ---
> fs/inode.c | 7 +++----
> 1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/fs/inode.c b/fs/inode.c
> index bd4da9c5207e..08d093737e8c 100644
> --- a/fs/inode.c
> +++ b/fs/inode.c
> @@ -192,8 +192,6 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
> inode->i_wb_frn_history = 0;
> #endif
>
> - if (security_inode_alloc(inode))
> - goto out;
> spin_lock_init(&inode->i_lock);
> lockdep_set_class(&inode->i_lock, &sb->s_type->i_lock_key);
>
> @@ -230,9 +228,10 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
> inode->i_flctx = NULL;
> this_cpu_inc(nr_inodes);
>
> + if (security_inode_alloc(inode))
> + return -ENOMEM;
This should probably be before this_cpu_inc(nr_inodes).
Powered by blists - more mailing lists